Biol Invasions DOI 10.1007/s10530-013-0459-2
ORIGINAL PAPER
A tritrophic approach to the preference–performance hypothesis involving an exotic and a native plant Taiadjana M. Fortuna • Jozef B. Woelke • Cornelis A. Hordijk • Jeroen J. Jansen • Nicole M. van Dam Louise E. M. Vet • Jeffrey A. Harvey
•
Received: 16 November 2012 / Accepted: 5 April 2013 Ó Springer Science+Business Media Dordrecht 2013
Abstract Exotic plants often generate physical and chemical changes in native plant communities where they become established. A major challenge is to understand how novel plants may affect trophic interactions in their new habitats, and how native herbivores and their natural enemies might respond to them. We compared the oviposition preference and offspring performance of the crucifer specialist, Pieris brassicae, on an exotic plant, Bunias orientalis, and on a related native plant, Sinapis arvensis. Additionally, we studied the response of the parasitoid, Cotesia glomerata to
Electronic supplementary material The online version of this article (doi:10.1007/s10530-013-0459-2) contains supplementary material, which is available to authorized users. T. M. Fortuna (&) J. B. Woelke C. A. Hordijk L. E. M. Vet J. A. Harvey Department of Terrestrial Ecology, Netherlands Institute of Ecology, P. O. Box 50, 6700 AB Wageningen, The Netherlands e-mail:
[email protected];
[email protected] J. J. Jansen Department of Analytical Chemistry, Institute for Molecules and Materials, Radboud University Nijmegen, P. O. Box 9010, 6500 GL Nijmegen, The Netherlands N. M. van Dam Department of Ecogenomics, Institute for Water and Wetland Research (IWWR), Radboud University Nijmegen, P. O. Box 9010, 6500 GL Nijmegen, The Netherlands
herbivore-induced plant volatiles (HIPV) and determined the volatile blend composition to elucidate which compound(s) might be involved in parasitoid attraction. On both host plants we also compared the parasitism rate of P. brassicae by C. glomerata. Female butterflies preferred to oviposit on the native plant and their offspring survival and performance was higher on the native plant compared to the exotic. Although, headspace analysis revealed qualitative and quantitative differences in the volatile blends of both plant species, C. glomerata did not discriminate between the HIPV blends in flight-tent bioassays. Nevertheless, parasitism rate of P. brassicae larvae was higher on the native plant under semi-field conditions. Overall, P. brassicae oviposition preference may be more influenced by bottom-up effects of the host plant on larval performance than by top-down pressure exerted by its parasitoid. The potential for dietary breadth expansion of P. brassicae to include the exotic B. orientalis and the role of top-down processes played by parasitoids in shaping herbivore host shifts are further discussed. Keywords Exotic invasive species Volatiles Plant preference–performance Host shift Multitrophic interactions Bunias orientalis
Introduction It has been long established that plants differ in their suitability for insect herbivores and in many species
123
T. M. Fortuna et al.
immature stages are often relatively immobile (Ehrlich and Raven 1964). Therefore, natural selection should favor female insects with an ability to discriminate between hosts of different suitability for the development of their progeny (Jaenike 1978; Mayhew 1997; Thompson 1988). The ‘preference– performance’ or ‘the mother-knows-best’ hypothesis suggests that females will maximize their fitness by laying eggs on plant species where their offspring will develop more successfully and thus have higher fitness returns (Jaenike 1978). Although some studies have shown a positive correlation between female preference and offspring performance (reviewed by Gripenberg et al. 2010; Singer et al. 1988), others report that this relationship is weak or is even absent altogether (Chew 1977; Larsson and Ekbom 1995; Ohsaki and Sato 1994; Rausher 1979; Valladares and Lawton 1991). Evaluating the factors influencing oviposition decisions in insects may be critically dependent on several parameters, such as the availability of high quality food plants in the habitat or dynamic changes in the physiological state of the foraging female (i.e. egg load, age, hunger level). In combination, these factors may make optimal foraging decisions in insects strongly context- and trait-dependent (Jaenike 1990; Papaj 2000). Several ecological factors have been proposed to explain apparent mismatches between choice and performance in insects (Mayhew 1997; Thompson 1988; Thompson and Pellmyr 1991). One of these factors is the recent addition of a novel plant species to the habitat. It has been documented that females may sometimes choose plant species that are of low quality and in some instances even fatal for the development of their offspring (Chew 1977; Keeler and Chew 2008; Ku¨hnle and Mu¨ller 2009; Larsson and Ekbom 1995). The inability of females to discriminate among plant species may occur if novel plants lack the required oviposition cues to reject them (Harvey and Fortuna 2012); or conversely, if these plants may possess oviposition stimulants, even when they are unsuitable as food plants (Renwick 2002). In this case, there may not have been sufficient time for females to evolve recognition cues that enable them to reject toxic novel plants or, when they do oviposit on these plants, for their progeny to adapt to the novel phytotoxins (Thompson 1988). These mechanisms are predicted by important hypotheses in invasion ecology, such as the enemy release (Keane and Crawley 2002) and the
123
novel weapons hypotheses (Callaway and Ridenour 2004), which have been proposed to explain the success of exotic invasive plant species in the new range. However, these mechanisms may be short lived and oviposition ‘‘mistakes’’ may in time also function as the ‘raw material’ for host plant shifts. Selection might favor females from populations that have broadened the number of host plant species in their diet, thereby saving time in searching for suitable host plants (Chew 1977; Graves and Shapiro 2003; Thompson and Pellmyr 1991). Another important factor that can explain the lack of correlation between female preference and larval performance is differential mortality risk imposed by pressures from natural enemies, such as predators or parasitoids (Price et al. 1980). Ohsaki and Sato (1994) showed that the differences in plant preference of three Pieris butterflies resulted from a trade-off between the avoidance of natural enemies (e.g. parasitoid wasps) and the intrinsic quality of several cruciferous food plants. Thus, although a plant species may be suboptimal nutritionally, it might also be unattractive to natural enemies of the herbivores and provide them with ‘enemy-free space’ (Price et al. 1980; Stamp 2001; Thompson 1988). Despite the many studies focused on the mechanisms underpinning female oviposition behavior and offspring performance (reviewed by Gripenberg et al. 2010), few have included the role of natural enemies in molding preference–performance relationships (Ohsaki and Sato 1994; Shiojiri et al. 2002), and even fewer have considered a tritrophic approach to plant preference– performance on exotic plant species (Harvey et al. 2010b). Resource exploitation by herbivores and their natural enemies occurs in habitats that are structurally and chemically heterogeneous. Individuals that use the information associated with the food sources (e.g. plant, prey or host) most efficiently will generally enjoy higher fitness benefits (Vet and Dicke 1992). Plant volatiles are often induced upon herbivore damage, and it is well established that natural enemies, such as parasitoids, use these volatile cues to locate host patches and hosts themselves (Vet and Dicke 1992). Although the functional and ecological importance of herbivore-induced plant volatiles (HIPVs) in parasitoid attraction has been discussed for many years (Gols et al. 2011; Takabayashi and Dicke 1996; Turlings et al. 1990), little is still known about the
A tritrophic approach to the preference–performance hypothesis
specificity of these plant compounds or mixtures of compounds in determining parasitoid host-finding behavior. The establishment and rapid spread of an exotic plant into a native community may fragment habitats and create barriers (e.g. chemical, physical) to the dispersal and host-finding abilities of herbivores and their natural enemies (Cronin and Haynes 2004; Harvey and Fortuna 2012). Furthermore, recent studies have shown that herbivore pressure was lower on exotic plant species than on native congeners, while potential top-down predator pressure was higher on exotic plants than on native ones (Engelkes et al. 2012). These findings suggest that lower herbivore loads on exotic species can also be related to high predation or parasitism rates on these species in their new range. In this study we compared plant preference–performance behavior of a specialist crucifer herbivore, Pieris brassicae L. (Lepidoptera: Pieridae), on two related wild species, the exotic Turkish rocket, Bunias orientalis L. (Capparales: Brassicaceae), and the native charlock mustard, Sinapis arvensis L. (Brassicales: Brassicaceae). Additionally, in flight chamber bioassays we compared the host-finding behavior of the P. brassicae endoparasitoid, Cotesia glomerata L. (Hymenoptera: Braconidae) on the two plant species, and measured parasitism rate under semi-field conditions outdoors. We addressed the following six questions: (1) Is oviposition choice of P. brassicae female butterflies different between the exotic and the native plant species? (2) Does the performance of their offspring differ between these food plants? (3) Do females prefer the host plant on which their offspring perform better? (4) Do the chemical profiles of HIPV emitted by the native and the exotic plant species differ? (5) Are C. glomerata wasps differentially attracted to the HIPV emitted by the native versus exotic plants? (6) Does this attraction result in different parasitism rates of P. brassicae? Based on our findings, we discuss the potential effects that exotic invasive plants may have on native trophic interactions.
Materials and methods Plants Bunias orientalis is a perennial species native to South-eastern Europe and West Asia that had been
established, albeit rare, in parts of Western Europe since the nineteenth century. In the 1980’s, it started to spread rapidly in many areas of its introduced range and it has become a highly invasive weed in several countries (Dietz et al. 1999; Harvey et al. 2010a). Previous studies have shown that this plant is unsuitable for the larval development of some crucifer specialists (Fortuna et al. 2012; Harvey et al. 2010a; Ku¨hnle and Mu¨ller 2009). However, females of the mustard leaf beetle, Phaedon cochleariae L., accepted B. orientalis for oviposition (Ku¨hnle and Mu¨ller 2009), and eggs and larvae of P. brassicae have been found on B. orientalis in the field in The Netherlands and Germany (T. Fortuna and J. Harvey, pers. observ.). Sinapis arvensis is an annual species and is a major food plant for crucifer specialists during part of the growing season in Europe (Feltwell 1982; Harvey et al. 2010a). Both plant species grow in similar ruderal habitats in dense stands. The native species was selected because its growing phenology, size and secondary chemistry (e.g. glucosinolates) are similar to B. orientalis (Harvey et al. 2010a). Insects The large white cabbage butterfly, Pieris brassicae, is a specialist herbivore, whose larvae feed exclusively on plants producing glucosinolates, including many species in the Brassicaceae family (Feltwell 1982). Wild broods of P. brassicae in early-mid larval stages were collected from several brassicaceous species growing along roadsides and cultivated fields in Gelderland in June 2009 and 2010. Caterpillars were then reared on Brussels sprouts plants, Brassica oleracea gemmifera cv. Cyrus, until they developed to pupal stage. After emergence, female and male butterflies were placed in an outdoor tent (2.5 9 2.0 9 2.0 m). Wild flowers of non-brassicaceous species were supplied as nectar sources. Naı¨ve mated female butterflies used in the oviposition choice experiment were 3–8 days old. Cotesia glomerata is a gregarious koinobiont endoparasitoid that attacks young larvae of several species of pierid butterflies; P. brassicae is its preferred host in Europe (Feltwell 1982). To locate their hosts, C. glomerata females rely on infochemicals (e.g. volatile cues) both from the plant and its host (Geervliet et al. 1997). Parasitoid wasps were obtained from parasitized P. brassicae caterpillars feeding on Brussels sprouts plants. Newly emerged adult wasps
123
T. M. Fortuna et al.
were placed in a rearing cage (30 9 30 9 30 cm) containing water and honey ad libitum. Females and males were placed together in a Petri dish with honey 1 day before the experiments, to ensure that female wasps were mated. For flight-tent and semi-field experiments we used naı¨ve-mated female wasps that were 5–10 days old when egg loads are fully matured (Jervis et al. 2008). All insects were reared in a controlled climate room (22 ± 2 °C, 60 % RH, 16L:8D). Experimental design Seeds of B. orientalis were collected from roadside populations growing in Gelderland (N52°000 E6°090 / N51°580 E5°420 /N51°580 E5°400 /N51°520 E6°000 ), and seeds of S. arvensis were collected from a large wild population in Vlieland (N53°170 E5°040 ), The Netherlands, in the summer 2008. B. orientalis seeds were extracted by cracking the seed coat as described by (Harvey et al. 2010a). Seeds were germinated in plastic containers (18 9 13 9 6 cm) with potting soil (30 % sand, 5 % clay and 65 % peat) and after 1 week, seedlings were transferred to 1.2 L pots and grown in a greenhouse (22/16 °C day/night, 60 % relative humidity (RH), 16L:8D photoperiod). Natural daylight was supplemented by 400 W metal halide lamps (225 lmol m-2 s-1 PAR). Plants were watered four times per week, and after 4 weeks, supplied with nutrient-enhanced Hoagland solution twice a week, to compensate for the nutrient depletion in the soil. Since the two plant species have slightly different growth rates, S. arvensis plants used in the experiments were 4–5 weeks old and B. orientalis plants were 6–8 weeks old, and not flowering. (1)
Plant-preference of Pieris brassicae in a semifield experiment
To determine differences in oviposition preference of wild P. brassicae female butterflies we conducted a dual-choice test. One female and two males were placed into an outdoor tent (2.5 9 2.0 9 2.0 m), containing six plants of each species (see experimental design in Figure S1, Supplemental Material). B. orientalis plants were placed on 20 cm stands to be at an equivalent height of S. arvensis plants. Wild flowers of non-brassicaceous species were provided in each tent as a nectar source for the butterflies. Female butterflies were allowed to oviposit for 24 h, and after the number of plants of each species with egg clutches
123
was counted. Tests were conducted during 2 summers, in June 2009 (25 replicates) and June 2010 (8 replicates). Fewer replicates were conducted in 2010 because of extremely hot weather conditions that killed many of the butterflies. Furthermore, we measured the leaf area of both plant species as an alternative factor influencing female oviposition choice. Leaf area of each plant species was determined using the software WinFOLIAÒ after scanning the surface area of all leaves of twelve plants chosen randomly from different tents. Statistical analysis All univariate analyses were run in GenStatÒ 14th edition (VSN International, Hemel Hempstead, UK). Butterflies oviposition choice data was analyzed using generalized linear mixed model (GLMM procedure Payne et al. 2011), with plant species effect as fixed factor and replicate (e.g. one female butterfly) as random effect. The binary trait ‘selected/unselected plants’ for oviposition followed a binomial distribution with a logit link function. Number of egg clutches in each chosen plant was modeled with generalized linear model (GLM) following a poisson distribution with a logarithm link function. Number of eggs per clutch and total leaf area of each plant species was tested using a two-sample t test (Payne et al. 2011). (2)
Performance greenhouse
of
Pieris
brassicae
in
the
Plants with eggs from the oviposition choice test conducted in 2010 were maintained in a greenhouse compartment. To ensure that differences in the offspring performance were due to differences in plant quality and not in genetic variation, we selected egg clutches from replicates where female butterflies oviposited on both plant species. Therefore, only two replicates out of eight were used in this experiment. Ten neonate larvae from hatched eggs were carefully transferred to a new plant and introduced with two extra plants in a cage (35 9 35 9 60 cm) according to the treatment. Four treatments were performed according to where the eggs were laid (oviposition plant) and where the larvae developed (larval development plant): (1) larvae hatched and reared on B. orientalis (BO-BO); (2) larvae hatched on B. orientalis and reared on S. arvensis (BO-SA); (3) larvae hatched and reared on S. arvensis (SA-SA); (4) larvae hatched on S. arvensis and reared on B. orientalis (SA-BO).
A tritrophic approach to the preference–performance hypothesis
Larval development was checked daily and new plants were supplied as required. Larval performance was measured based on survival rate, pupal weight and development time. Survival was determined as the fraction of larvae that survived until adult stage from the total number that was placed on the plants. Day of pupation was recorded when pupae were formed and within 24 h, when the cuticle had hardened, pupae were removed from the cages and weighed on an analytical balance (Mettler-Toledo AG104, accuracy ±1 mg). After weighing, pupae were individually placed in plastic vials with a perforated lid until adults emerged. Development time was then recorded as the number of days from egg hatching until adult emergence. BO-BO and BO-SA treatments were replicated 10 times and SA-SA and SA-BO treatments 8 times. Statistical analysis Performance data of butterflies’ offspring was analyzed using linear mixed model (restricted maximum likelihood method, Payne et al. 2011) with host plant, oviposition and development plant, and their interaction as fixed factor and maternal effect as random factor. In all parameters of performance, the interaction term was never significant, so that was removed from the model. Prior to analyses, survival data was arcsine square-root transformed to meet the assumption of normality. Pupal weight and egg-to-adult development time data was averaged within replicate. (3)
Plant-preference of Cotesia glomerata in a flighttent experiment
The flight response of C. glomerata females to both plants was tested in a white nylon sheeting tent (2.5 9 2.0 9 2.0 m) that was placed in the greenhouse (22 ± 0.5 °C, 50 % RH). Two plants were placed, according to the test (see below), on a table in the middle of the tent, 5 min before the bioassay. One day before testing, naı¨ve wasps were collected from the mass-rearing and placed individually in 5-mL vials with honey and water. Wasps were then released from the vials at the other end of the table with the experimental plants (see Figure S2, Supporting Information), and because wasps do not leave the vials immediately, 10 wasps were released simultaneously. Wasps were allowed to fly freely inside the tent and as soon a wasp alighted on a plant, it was collected and the plant species on which it had landed was recorded. Additional wasps were released when the previous
wasps exhibited no flight response within 15 min. A total of 10 wasp responses were recorded per each set of plants (replicate). Plant position was changed between replicates. Individual wasps were used only once. Test 1 Undamaged (control) versus Damaged B. orientalis (DB20). Herbivore damage was inflicted by 20 L2 P. brassicae larvae that were carefully placed on a clean leaf and allowed to feed for 48 h. To enhance the HIPV emission, damaged plants were presented to the wasps containing host larvae, which habitually fed gregariously on the underside of a leaf. Eight replicates were conducted. Test 2 Damaged B. orientalis (DB20) versus Damaged S. arvensis (DS20). Plants of the two species were infested by 20 L2 P. brassicae larvae for 48 h. B. orientalis plants were placed on 10 cm stands to be at an equivalent height with S. arvensis plants. Sixteen replicates were conducted. Test 3 Higher host-density on Damaged B. orientalis (DB40) versus Damaged S. arvensis (DS20). B. orientalis plants were infested by 40 L2 P. brassicae larvae while S. arvensis plants were infested by 20 L2 P. brassicae larvae for 48 h. Thirteen replicates were conducted. Moreover, the feeding damage of P. brassicae larvae on both plant species was determined a posteriori with the software WinFOLIAÒ after scanning the damaged area of the tested plants. Statistical analysis Parasitoid plant preference data was analyzed using GLMM with plant species effect as a fixed factor and replicate (e.g. 10 wasp responses) as a random factor. These data followed a binomial distribution with a logit link function. (4)
Headspace analysis of plant volatiles
Volatile organic compounds (VOC) were collected from B. orientalis (n = 9) and S. arvensis (n = 10) plants that had been undamaged or damaged by L2 P. brassicae larvae for 48 h. Potted plants were transferred individually to a 17-L glass collection container (41 9 24.5 cm) and placed in a controlled climate cabinet (21 ± 0.5 °C, 70 % RH). At the top, the containers were supplied constantly with 200 mL pressurized air and were cleaned over a Zero Air generator to remove hydrocarbons (Parker Hannifin
123
T. M. Fortuna et al.
Corp, Tewksbury, USA). Volatiles were trapped on 150 mg Tenax TA and 150 mg Carbopack B steel trap (Markes, Llantrisant, UK) for 60 min at a flow of 100 mL min-1. Volatiles were desorbed from the Tenax traps using an automated thermodesorption cold trap unit (model Unity, Markes) at 200 °C for 12 min (He from 30 mL min-1). The released compounds were cryofocused in an electrically cooled (-10 °C) sorbent trap to allow for narrow starting bands. Volatiles were injected into GC in sliptless mode by ballistic heating of the cold trap for 3 min to 270 °C. After separation by capillary gas chromatography (column: 30 m 9 0.32 mm internal diameter RTX-5 Silms, film thickness 0.33 lm; 40–95 °C at 3 °C min-1, then to 165 °C at 2 °C min-1, and after to 250 °C at 15 °C min-1), volatiles were directly introduced in mass spectrometer (MS; model Trace, ThermoFinnigan, Austin, USA) operating at 70 eV in EI ionization mode. Mass spectra were recorded with full scan mode (33–300 AMU, 3 scans s-1). Compounds were identified by comparing their mass spectra with those in reference manuals (Adams 2007), National Institute of Standards and Technology (NIST, USA 2008; http://www.nist.gov), Wiley 7th edition spectral libraries, and by checking the retention indices. For identification and quantification we used AMDIS 2.1 (Automated Mass spectral Deconvolution and Identification Software). Individual compounds were quantified by measuring peak areas of the total ion fractions of the integrated signal. Peak areas were divided by the total volume (mL) that was sampled over each trap to correct for small differences in sampling time and flow rates over individual traps. To avoid rare compounds of having disproportionate effects on volatile profiles, only compounds with peak area larger than the background and found in more than 80 % of the plants in at least one treatment group were analyzed. Finally, peaks of impurities were removed, which resulted in a dataset of 63 volatiles in 38 plants. Statistical analysis Differences in VOC profiles among treatments were analyzed using a supervised multivariate analysis method, orthogonal partial least squares discriminant analysis (OPLS-DA). This method allows us to describe the differences between plant groups in each treatment by comparing with other unrelated chemical variation (Bylesjo et al. 2006). As treatment pairs, we used the three comparisons of the parasitoid choice tests: (1) undamaged and
123
damaged B. orientalis plants (control vs. DB20); (2) damaged B. orientalis and S. arvensis plants with the same host density (DS20 vs. DB20); (3) or with different host density (DS20 vs. DB40). These comparisons were analyzed using three separate OPLSDA models and the optimal number of latent variables for each model was evaluated by cross-validation following the procedure described in (Pierre et al. 2011). The VOC discriminating most between two contrasted plant groups in each comparison were collected in a weight vector, explaining the relative importance of each volatile within the total blend. The weight vectors of all three comparisons contained the ranks of the same VOCs, which were represented in a two-dimensional bubble plot to help visualization (Fig. 4). Based on the position and size of the bubble of each compound together with an arbitrary threshold (0.05) for the weight vectors (Fig. 4), we defined potential bioactive compounds as candidates responsible for the parasitoid behavior in the bioassays (Fig. 3). Multivariate analysis were performed using MATLAB version 7.12 (R2011a; MathWorks, Natick, Ma, USA). (5)
Host-acceptance of Cotesia glomerata in a semifield experiment
Parasitoid host location and acceptance behavior was studied in a dual-choice test performed in an outdoor tent (2.5 9 2.0 9 2.0 m), containing six plants of each species (see experimental design in Figure S1, Supplemental Material). Prior to the test, ten L2 P. brassicae larvae were placed on the plants and allowed to feed for 24 h. Additionally, naı¨ve female wasps of C. glomerata from the mass-rearing were separated in groups of five and placed in 5-mL vial with honey and water. On the test day, one vial was released per tent and wasps were allowed to search for their hosts for 4 h. In each tent, 120 P. brassicae hosts were presented to C. glomerata wasps. The test was replicated nine times and 45 wasps were released in total. After testing, wasps were collected and plants with P. brassicae caterpillars were brought to the greenhouse, where they fed for 4 days to enable the parasitoid development from eggs to larvae. Caterpillars from each plant were placed in 2-mL Eppendorf tubes and kept in the freezer at -20 °C until dissection under a stereomicroscope (Leica M205 C). The primary clutch size (number of
A tritrophic approach to the preference–performance hypothesis
eggs and larvae inside the host) of C. glomerata was recorded and parasitism rate was determined as the proportion of parasitized caterpillars over the total number of caterpillars collected from each plant after the choice test. Statistical analysis Percentage data on herbivore parasitism rate was analyzed using Chi Squared test to compare the proportion of parasitized caterpillars on each plant species. Total herbivore leaf damage and parasitoid primary clutch size was tested using a twosample t test (Payne et al. 2011). Prior to analysis, wasp clutch size was summed within individual plant and averaged per plant species within replicate. Chi Squared test that was performed in PopTools (Microsoft Excel add-in; Hood 2010).
(a)
(b)
Results (1)
Plant-preference of Pieris brassicae in a semifield experiment
Female butterflies significantly preferred to oviposit on plant of S. arvensis over B. orientalis in both years that the test was performed (2009: W = 12.92; P \ 0.001; 2010: W = 5.09; P = 0.024; Fig. 1a). Overall, 85 % of the butterflies that exhibited preference chose the native over the exotic plant. In 2010, despite extremely hot conditions that occurred during the choice test, the butterflies still showed a similar oviposition response as in the previous year. Nevertheless, some P. brassicae butterflies (39 %) laid their eggs on both plant species. Host plant species had no significant effect on the number of egg clutches laid by the butterflies in both years (2009: Z = -0.05; P = 0.956; 2010: Z = -0.71; P = 0.477). Generally, butterflies laid only a single clutch per plant, independent of the plant species. However, the number of eggs laid per clutch was significantly higher on the exotic plant compared to the native plant (t = -2.16; P = 0.039) in 2009 (Fig. 1b). The average clutch size on B. orientalis was approximately 90 ± 14 eggs, whereas on S. arvensis was 60 ± 7 eggs. In 2010, although the average clutch size was larger on the exotic plant (35 ± 17 eggs) compared to the native plant (15 ± 3 eggs) no significant difference was detected (t = -1.19; P = 0.277; Fig. 1b). Total leaf area varied significantly according to the plant species
Fig. 1 Oviposition behavior of wild Pieris brassicae butterflies on Sinapis arvensis (white bar) and Bunias orientalis (dark grey bar) plants. Females were allowed to lay eggs in a two-choice test for 24 h. Butterflies were tested individually in the summer of 2009 and 2010. Given is a Mean ± SE number of plants preferred for oviposition; b Mean ± SE number of eggs laid per clutch. Asterisks indicate significant difference in the number of plants with eggs (Wald test: P \ 0.001**; P \ 0.05*) and in the number of eggs per clutch (two-sample t test: P \ 0.05*). n = 25 (2009); n = 8 (2010)
(t = 6.11; P \ 0.001). The exotic plant had a leaf area (1424 ± 107 cm2) that was almost twice as large as that of the native plant (741 ± 35 cm2). (2)
Performance greenhouse
of
Pieris
brassicae
in
the
Survival of P. brassicae offspring was significantly affected by the plant on which butterflies oviposited their eggs (W = 11.49; P \ 0.001), but not on the plant on which their larvae grew (W = 0.08; P = 0.777). Independent of the plant where the herbivore developed, survival was higher in clutches laid on S. arvensis, 96 %, (SA-SA; SA-BO) than on B.
123
T. M. Fortuna et al.
orientalis, 78 %, (BO-BO; BO-SA; Fig. 2a). Pupal weight was significantly affected by the oviposition plant (W = 4.99; P = 0.033) and larval development plant (W = 178.9; P \ 0.001). Pupae developing on S. arvensis plants were bigger when they were from S. arvensis-preferring females (SA-SA) than those from B. orientalis-preferring females (BO-SA). The same pattern was found on B. orientalis plants, where pupae from S. arvensis-preferring females (SA-BO) were larger than those from B. orientalis-preferring females (BO-BO; Fig. 2b). Egg-to-adult development time was also significantly affected by the oviposition plant (W = 4.78; P = 0.036) and the larval development plant (W = 43.40; P \ 0.001). Caterpillars developed faster on the native plant (SA-SA; BOSA) than on the exotic plant (BO-BO; SA-BO) (Fig. 2c). (3)
(b)
Parasitoid plant-preference in a flight-tent experiment
C. glomerata showed a clear preference for B. orientalis plants damaged by the larval host, when compared with undamaged plants (F = 68.15; P \ 0.001). Almost 90 % of the wasps landed on plants that were infested with P. brassicae larvae, confirming that the parasitoid females are able to locate their hosts on the exotic plant (Fig. 3a). However, the parasitoid did not discriminate between host-infested plants of S. arvensis and B. orientalis when both plant species were infested with the same host density (F = 0.80; P = 0.379; Fig. 3b). Female wasps only showed a preference for the exotic plant when it was infested with higher host density (F = 28.58; P \ 0.001). Significantly more wasps landed on B. orientalis plants that were infested with 40 caterpillars compared to S. arvensis plants infested with 20 caterpillars (Fig. 3c). The total leaf area damaged by 20 P. brassicae caterpillars was significantly greater in S. arvensis plants than in B. orientalis plants with the same host density (t = -2.26; P = 0.038), but smaller than in B. orientalis plants with 40 caterpillars (t = 4.85; P \ 0.001). (4)
(a)
Headspace analysis of plant volatiles
A total of 63 volatile compounds were detected in the headspace analysis of B. orientalis and S. arvensis plants (Table S1, Supporting Information). In both species, control and herbivore-infested plants differed quantitatively in their VOC emissions. However,
123
(c)
Fig. 2 Survival (a), pupal weight (b) and egg-to-adult development time (c) of Pieris brassicae offspring grown on Sinapis arvensis (SA) or Bunias orientalis (BO) plants. Four plant treatments are considered according to the oviposition plant (first two letters SA- or BO-) and larval development plant (last two letters -SA or -BO). Given is mean ± SE for each plant treatment. Significant differences are shown for oviposition plant and larval development plant (Wald test: P \ 0.001; P \ 0.05). n = 8 (SA-SA, SA-BO); n = 10 (BO–BO, BO-SA)
while in B. orientalis there was threefold increase of VOC in damaged plants compared to control, in S. arvensis there was a twofold decrease of VOC in damaged plants. Despite the total HIPV decrease in S. arvensis damaged plants, the concentration of some compounds increased upon herbivory (see Table S1 in Supporting Information). HIPV profile in B. orientalis
A tritrophic approach to the preference–performance hypothesis
Fig. 3 Percentage of choices of Cotesia glomerata females, in a two-choice experiment between: a undamaged (control) or damaged Bunias orientalis plants by 20 Pieris brassicae caterpillars (DB20); b Sinapis arvensis (DS20) or B. orientalis plants (DB20) both damaged by 20 P. brassicae caterpillars; c S. arvensis plants damaged by 20 P. brassicae caterpillars (DS20) or B. orientalis plants damaged by 40 P. brassicae caterpillars (DB40). Double asterisk indicate significant difference in the parasitoid preference tests (F test: P \ 0.001). a n = 8; b n = 16; c n = 13
plants also varied quantitatively among treatments. Volatile emissions were higher in plants damaged by large herbivore load. Between plant species, HIPV profiles were both qualitatively and quantitatively different; damaged plants of B. orientalis produced higher compound variety and at a three time higher rate than damaged plants of S. arvensis. The computation of OPLS-DA based on VOC emission showed that the effect of all three comparisons from the parasitoid behavior bioassay can be described by low-dimensional multivariate models. The models consisted either in two latent variables, for the first comparison (control vs. DB20), or in one latent variable, for the second and third comparisons (DS20 vs. DB40; DS20 vs. DB20). All models had a significant permutation-based probability (P = 0.02; P \ 0.01; P \ 0.01, respectively), supporting that each tested comparison is real. Differences and similarities among plant treatments were compared by plotting the discriminate functions of the three OPLS-DA models against each other on a bubble plot (Fig. 4). This representation of OPLS-DA in twodimensions allows, therefore, visual identification of interesting compounds that are important for differentiation between the volatile profiles for each of the plant treatments. Particularly, decanal (9),
cyclohexane, decyl- (19), ethyl hexanoate (32), ethanone, 2-hydroxy (39), 2-butanone (44), menthone isomer (53), unknown terpene (61), polyalcene (62) were in higher concentration in damaged plants of B. orientalis compared to control (X-axis) or to damaged plants of S. arvensis (Y-axis; Fig. 4). By contrast, octanol (1), 1-butanol 3-methyl- (3), methanols, benzyl- (5), menthol isomer (7), 4-nonenal (15), anthracene (57) were in the reverse situation (Fig. 4). Combining this information with the position and size of the compound bubbles we looked for potential bioactive compounds responsible for the parasitoid behavior in the bioassays (Fig. 3). The volatile candidates have: (1) to increase or decrease similarly in DB20 compared to control, and in DB40 compared to DS20 (upper right or lower left quarter); (2) to show no differences (in- or decrease) between DB20 and DS20 (large bubble size; Fig. 4). A relevant pattern for potential bioactive compounds was followed by p-Menthane (24). Additionally, a similar behavior is followed by the compound Ethanone, 2-hydroxy-1,2diphenyl (39) and Methanol, benzyl- (5), although considering a slightly higher threshold level (0.1) (Fig. 4; Table S1 Supporting Information). (5)
Host-acceptance of Cotesia glomerata in a semifield experiment
Parasitism rate by C. glomerata differed significantly on the two plant species (v2 = 9.25; P = 0.002). Female wasps parasitized a higher number of P. brassicae larvae on the native plant than on the exotic plant (Fig. 5). However, parasitoid primary clutch size was not significantly different between hosts developing on the two plant species (t = 0.11; P = 0.916). Female wasps typically laid primary broods of 22–25 eggs in P. brassicae caterpillars on the two species.
Discussion Host-plant preference–performance of Pieris brassicae Our results show a clear preference–performance link in the specialist herbivore, P. brassicae. The butterflies showed an innate oviposition preference for the native host plant, S. arvensis, over the exotic species, B. orientalis, in both years the test was conducted.
123
T. M. Fortuna et al.
Fig. 4 Two-dimensional bubble plot representation of volatiles emitted after different herbivore infestation treatment in Bunias orientalis and Sinapis arvensis plants. Herbivory was induced using 20 or 40 P. brassicae larvae. Plant treatments were compared by plotting the discriminate functions of the three OPLS-DA models against each other. (1) First comparison (X-axis) represents the VOC profile emitted by damaged plants of B. orientalis in contrast to control plants (control vs. DB20). (2) Second comparison (Y-axis) corresponds to HIPV emitted by damaged plants of S. arvensis and B. orientalis with different host densities (DS20 vs. DB40). (3) Third comparison (bubble size) corresponds to HIPV emitted by damaged plants of S. arvensis and B. orientalis with the same host density (DS20 vs. DB20). Each bubble represents a volatile compound and its position is determined according to its relative importance (weight vector values) for each of the three comparisons. Increased or decreased levels of a compound are indicated by a positive or negative value in X- and Y-axis (1 and 2), or by the size of the circles (3). The size of the bubble is negatively correlated with the relative importance of the compound in explaining differences in the comparison (3) i.e. large circle size indicates a compound with relatively low importance in explaining volatile differences between DS20 and DB20 treatments. Numbers correspond to the compounds (referenced in Table S1) with relative importance (-0.05 \ weight vector [ 0.05) in the first and second comparison. An arbitrary threshold of significance was defined by an absolute weight vector value smaller than 0.05 (grey bubble). The black bubble (24) corresponds to the biomarker p-Menthane. B. orientalis: n = 9; S. arvensis n = 10
Furthermore, the plant species on which they oviposited affected the survival and development of their offspring. Overall, caterpillars survived better when females oviposited on the native plant, and pupae that had developed on this plant species were larger and completed their development faster, than those which had developed on the exotic plant. These results
123
Fig. 5 Parasitism rate (%) of Cotesia glomerata females in a two-choice experiment between Sinapis arvensis (white bar) and Bunias orientalis (dark grey bar) plants infested with Pieris brassicae larval hosts. Plants were infested with 10 P. brassicae caterpillars 24 h prior the choice test. Asterisk indicates significant difference (v2 test: P \ 0.05). n = 9
emphasize the importance of oviposition decisions made by the female butterflies and provide further support for the ‘mother knows best’ hypothesis (Jaenike 1978). Furthermore, our study supports the results of Gripenberg et al. (2010) who found stronger preference–performance relationships in oligophagous insect herbivores than in either monophagous or polyphagous insect herbivores. Indeed, a positive correlation between female plant choice and offspring performance has been demonstrated in several studies with other oligophagous insects (Ku¨hnle and Mu¨ller 2009; Scheirs et al. 2003; Travers-Martin and Mu¨ller 2008). For instance, plant preference in females of the sawfly Athalia rosae, a specialist herbivore of brassicaceous species, was influenced by larval performance, and reproductive success was found to be highly dependent on host plant choice over the course of only two generations (Travers-Martin and Mu¨ller 2008). In an earlier study, Harvey et al. (2010a) also showed that P. brassicae preferred to oviposit on the native crucifers, Brassica nigra and S. arvensis, over B. orientalis. However, the survival of this herbivore, which had been reared for many generations on cultivated cabbages in the laboratory, was much lower on B. orientalis compared to the results obtained in the present study (Fortuna et al. 2012; Harvey et al. 2010a). These partly contrasting results suggest that long-term inbreeding in lab-reared insects may reduce
A tritrophic approach to the preference–performance hypothesis
their vigor and make them more susceptible to novel plant species. Laboratory cultures may lose much of the genetic variability related to behavioral and physiological plasticity of insects in natural ecosystems (Harvey et al. 2010a). Thus, greater efforts should be made in using wild individuals in these bioassays to account for the natural variation among the insect populations and also because they may possess a greater ability to deal with chemical defenses of natural plants (Harvey et al. 2010a).
Host preference of Cotesia glomerata In the flight-tent bioassays, C. glomerata females detected volatile blends of B. orientalis plants damaged by P. brassicae hosts, both when they were tested against undamaged plants and against S. arvensis damaged by lower host density. However, female wasps did not discriminate between B. orientalis and S. arvensis with the same host density. This behavior is common to parasitoid species whose hosts are specialists but that feed on different plant species within the same family (Vet and Dicke 1992). As a consequence, different parasitoid generations can also be associated with different host-associated plants within a growing season (Gols et al. 2011). In these cases, natural selection may favor a flexible response of naı¨ve parasitoids to plant infochemicals due to the variation among blends emitted by different plant species. Moreover, to effectively expand their hostforaging arena, parasitoids may actively choose not to discriminate between subtle differences until they become more sensitized to these differences through experience (Vet et al. 1998). However, despite similar attraction to the host plant cues, in our semi-field experiment the wasps parasitized slightly more P. brassicae larvae on S. arvensis than on B. orientalis plants. This may be due to differences in the spatial scale of the two experiments, which result in a different localized perception of the parasitoids to plant- and host-related cues. During host plant selection behavior C. glomerata rely more on general plant cues (e.g. HIPV) common to the Brassicaceae family, whilst during the host selection phase the parasitoid also uses contact cues based partly on host qualityrelated parameters, such as host size and nutritional status (Vinson 1976). Additionally, other plant-related traits such as height and visibility may have also
influenced differently the parasitoid host searching behavior in the two experimental bioassays. Many parasitoids, including C. glomerata, are known to have evolved highly efficient mechanisms to perceive chemical cues that are associated with high quality hosts (Geervliet et al. 1998). Moreover, our study highlights the importance of validating the results obtained in flight-chamber bioassays with field or semi-field experiments. Plant choice experiments in parasitoids have mostly been conducted in y-tubes (Bukovinszky et al. 2005) and flight assays (Geervliet et al. 1996, 1998; Gols et al. 2008), which only tell us where the wasps alight, but not whether they actually locate and accept hosts on the tested plants. Field experiments can indicate whether and to what extent behavioral and developmental parameters measured in the laboratory relate to conditions that occur in more complex situations in natural habitats (Geervliet et al. 2000). In addition, it has been suggested that flightchambers enclosed environment may lead to less contrasting odor-sources or volatile interference, which could also affect plant preference by parasitoids (Vos et al. 2001).
Role of the volatiles in Cotesia glomerata attraction The volatile bouquet emitted by P. brassicae-damaged plants of B. orientalis was more diverse and concentrated than that emitted by S. arvensis plants. These results agree with previous studies, where volatile emissions varied considerably in quality and quantity, both within and among Brassicaceous species (Bukovinszky et al. 2005; Gols et al. 2011; Gols et al. 2008). However, despite the qualitative and quantitative differences in HIPV emitted by the two plant species, C. glomerata did not discriminate between these plants during its host-searching behavior. Similar results were found in the studies of Geervliet et al. (1996, 1997) where C. glomerata females did not discriminate between different plant cultivars infested with hosts, even when the volatile profiles of those cultivars were found to be qualitatively different. Nevertheless, results from OPLS-DA models showed that the cyclic alkane p-menthane might be a good candidate to explain the parasitoid behavioral response in the flight-tent bioassays. This alkane is known to be involved in terpenoid biosynthesis as a
123
T. M. Fortuna et al.
parent structure of monoterpenes (Bernards 2010). For example, it has been demonstrated in the menthol biosynthesis pathway of Mentha species that p-menthane is the precursor of the monoterpene limonene, which is in turn the precursor of menthone (Croteau et al. 2005). Also, in our system a menthone isomer is emitted by both plant species, which can be related to a similar biosynthesis pathway, where p-menthane is used as a terpenoid precursor. Monoterpenes, due to their volatile character, are known to have important biological activities as aroma components, and they can mediate tritrophic interactions acting as parasitoid wasp attractants (Mumm and Dicke 2010). It has already been shown that C. glomerata females have the ability to detect terpenes with their antennal chemical receptors (Smid et al. 2002; Vet et al. 1995). In addition, this parasitoid species seems to rely on more general plant cues associated with its host. Therefore, it is likely that female wasps have evolved to detect terpenoid precursors, because they are more common and less variable than the compounds at the end of the terpene biosynthetic pathway. Moreover, these precursor compounds are released during an earlier stage of host infestation, which might indicate more suitable host stages for the parasitoid. In spite of numerous factors affecting parasitoid host location, in particular the role played by HIPVs, it is still unknown how parasitoids perceive and integrate shifts in the composition of complex VOC mixtures, and how they learn to respond to these changes within complex odour plumes (Dicke and Vet 1999; Harvey and Fortuna 2012). For instance, herbivorous spider mite Tetranychus urticae and its specialist enemy mite, Phytoseiulus persimilis were shown to perceive volatile mixtures as a whole rather than as a collection of individual components (van Wijk et al. 2011). Further investigations based on gas chromatography coupled to electroantennography are required to reveal if compounds, such as the ones identified in the present study, elicit a response in the chemoreceptors of the parasitoid (Dicke and Baldwin 2010). Using this technique, Smid et al. (2002) identified the volatiles emitted by P. brassicae-damaged cabbage plants to which the parasitoid evoked EAG reaction in its antennae. Among those volatile compounds, four were found in the present study from either B. orientalis or S. arvensis plants (e.g. 2hexenal, octanal, limonene, decanal) and although they seem to not explain the parasitoid choice based on
123
OPLS-DA, they are likely building stones of the whole odour used by the parasitoids. Ecological significance In their new range exotic plant species can disrupt trophic interactions and food webs through physical and chemical interference as well as by displacing native plants (Cronin and Haynes 2004; Harvey and Fortuna 2012). However, in time they may ultimately generate new communities to which local native consumers may adapt. This is particularly the case if the exotic plants are closely related to the natives and thus produce chemicals to which the native herbivores are already pre-adapted (Harvey and Fortuna 2012). As we showed, some P. brassicae butterflies oviposited on B. orientalis plants in our semi-field experiment and the survival of their progeny was relatively high on this plant. On the other hand, as predicted by the enemy release hypothesis other native plant species, such as S. arvensis, are still likely to be highly preferred, because of long co-evolutionary history between the butterflies and these ubiquitous native plants. This may end up reducing the rate at which the novel plant will be used as an oviposition site. However, P. brassicae and other crucifer specialist herbivores may benefit by incorporating B. orientalis into their diet as an alternative host plant, particularly in late summer when other crucifers are scarce. Acceptance of a new plant species for oviposition by a female is an important first step towards dietary expansion. It has been argued that in general behavioral plasticity (such as oviposition preference) should undergo more rapid evolutionary changes than morphological or physiological plasticity (such larval growth on different plants), and that for this reason behavioral plasticity will be more likely to be an initiator of new direction in evolution (West-Eberhard 1989). Hence, the ovipositing females can take the leading role in the evolution of host plant utilization. For example, in the garlic mustard-Pieris oleracea association in United States, the butterflies readily accept this invasive plant for oviposition, although their larvae develop very poorly on it (Keeler and Chew 2008). However, more recently, in habitats where the garlic mustard has been well established for few decades and is the dominant crucifer, larval survivorship was shown to be positively correlated with the mother’s preference, suggesting that
A tritrophic approach to the preference–performance hypothesis
P. oleracea is adapting to this plant (Keeler and Chew 2008). Similarly, P. brassicae may also gain the advantage of incorporating B. orientalis into its diet if this plant becomes more abundant, which is occurring in parts of central and northern Europe. In this scenario, native specialist herbivores would be positively selected to adapt to the novel food plant (Chew 1977). In the Netherlands, P. brassicae can have two (or three) generations within the growing season. Given that the native annual crucifers also have very short life cycles (i.e. typically 1–2 months) and varying seasonal phenologies, different generations of this butterfly probably develop on different plant species (Gols et al. 2011). B. orientalis grows from April until October, thus different generations of the butterfly may have the ability for developing on the same species over an entire growing season. The progeny of a female can remain in the natal patch, without having to spend potentially costly time searching for new plants in new habitats. In this instance, a preference–performance correlation may be produced within a single generation, because individuals that survive on a particular host species will tend to pass on preference–performance genes to their progeny (Singer et al. 1988). Additionally, individual B. orientalis plants often produce enough leaf biomass to support a large brood of P. brassicae, whereas annual crucifers, like S. arvensis, rarely do (J Harvey, pers. obs.). In addition, the evolution of habitat preference may also be driven by top-down pressure mediated by natural enemies. For instance, enemy-pressure may be greater when herbivores develop on some plant species, and thus selection of certain plant types enables the herbivores to escape their natural enemies, i.e. enemy-free space (Camara 1997; Stamp 2001; Thompson 1988). Bunias orientalis may create an enemy-free space for P. brassicae if their larvae suffer lower parasitism rates from C. glomerata than larvae developing on native species. In some herbivore species it has been shown that the evolution of plant preference can be driven by both intrinsic quality of food plants and interactions with predator and/or parasite loads. In other pierids, plant preference may be based on trade-offs, whereby butterflies choose nutritionally sub-optimal plants but where their progeny are less susceptible to attack by parasitoids (Ohsaki and Sato 1994).
By affecting the foraging efficiency of natural enemies, such as predators and parasitoids, plants can affect the impact that those enemies have on herbivore populations and thus ultimately interfere with predator–prey or parasitoid-host dynamics (Dicke and Vet 1999). Parasitoids can increase their efficiency in host finding by learning plant-related cues (e.g. HIPV) and temporarily specialize on available and profitable plants (Geervliet et al. 2000). Cotesia glomerata, for example, learns to prefer plant species that are most profitable in terms of host encounter rate (Geervliet et al. 1998). By delaying herbivore development, B. orientalis increases the chance that the parasitoid encounters the more suitable younger host stages (the slow-growth-high-mortality hypothesis, Benrey and Denno 1997). Additionally, since the butterflies tend to lay larger broods on the exotic plant, this may also increase HIPV emission, leading to greater parasitoid attraction (Vet et al. 1998), as observed in our bioassay. These factors may allow C. glomerata to easily locate its P. brassicae host on the novel B. orientalis with which it may have no co-evolved relationship. Therefore, the few herbivores that may realize host shifts to the novel plant species can be subjected to higher parasitism pressure, and ultimately release the exotic plants from local herbivores in the invaded environment. Importantly, exotic plants offer an exciting opportunity to investigate the evolutionary forces involved in potential host shifts of herbivores and their natural enemies, ultimately providing insight into some of the many factors that influence the success of novel plants in their new range. Acknowledgments We thank Roel Wagenaar for rearing the parasitoid wasps; Leo Westerd for supplying butterfly eggs; Gregor Disveld for the technical assistance in the greenhouse; Olga Kostenko and Koen Verhoeven for the fruitful discussions on the statistical analyses; Keith Clay, Antoine Branca, and two anonymous reviewers for very valuable suggestions; This work was funded by a PhD-fellowship from the Portuguese governmental institution, Fundac¸a˜o para a Cieˆncia e Tecnologia, to T. M. Fortuna (SFRH/BD/40531/2007).
References Adams RP (2007) Identification of essential oil components by gas chromatography/mass spectrometry. Allured Publishing Corporation Benrey B, Denno RF (1997) The slow-growth-high-mortality hypothesis: a test using the cabbage butterfly. Ecology 78:987–999
123
T. M. Fortuna et al. Bernards MA (2010) Plant natural products: a primer. Can J Zool 88:601–614 Bukovinszky T, Gols R, Posthumus MA, Vet LEM, van Lenteren JC (2005) Variation in plant volatiles and attraction of the parasitoid Diadegma semiclausum (Hellen). J Chem Ecol 31:461–480 Bylesjo M, Rantalainen M, Cloarec O, Nicholson JK, Holmes E, Trygg J (2006) OPLS discriminant analysis: combining the strengths of PLS-DA and SIMCA classification. J Chemometr 20:341–351 Callaway RM, Ridenour WM (2004) Novel weapons: invasive success and the evolution of increased competitive ability. Front Ecol Environ 2:436–443 Camara MD (1997) A recent host range expansion in Junonia coenia Hubner (Nymphalidae): oviposition preference, survival, growth, and chemical defense. Evolution 51:873– 884 Chew FS (1977) Coevolution of pierid butterflies and their cruciferous foodplants. II. The distribution of eggs on potential foodplants. Evolution 31:568–579 Cronin JT, Haynes KJ (2004) An invasive plant promotes unstable host-parasitoid patch dynamics. Ecology 85:2772–2782 Croteau RB, Davis EM, Ringer KL, Wildung MR (2005) (-)Menthol biosynthesis and molecular genetics. Naturwissenschaften 92:562–577 Dicke M, Baldwin IT (2010) The evolutionary context for herbivore-induced plant volatiles: beyond the ‘cry for help’. Trends Plant Sci 15:167–175 Dicke M, Vet LEM (1999) Plant-carnivore interactions: evolutionary and ecological consequences for plant, herbivore and carnivore. In: Olff H, Brown VK, Drent RH (eds) Herbivores between plants and predators. Blackwell Science, London, pp 483–520 Dietz H, Steinlein T, Ullmann I (1999) Establishment of the invasive perennial herb Bunias orientalis L.: an experimental approach. Acta Oecol 20:621–632 Ehrlich PR, Raven PH (1964) Butterflies and plants: a study in coevolution. Evolution 18:586–608 Engelkes T, Wouters B, Bezemer TM, Harvey JA, van der Putten WH (2012) Contrasting patterns of herbivore and predator pressure on invasive and native plants. Basic Appl Ecol 13:725–734 Feltwell J (1982) Large white butterfly: the biology, biochemistry and physiology of Pieris brassicae (Linnaeus). Dr W. Junk Publishers, The Hague Fortuna TM, Vet LEM, Harvey JA (2012) Effects of an invasive plant on the performance of two parasitoids with different host exploitation strategies. Biol Control 62:213–220 Geervliet JBF, Vet LEM, Dicke M (1996) Innate responses of the parasitoids Cotesia glomerata and C. rubecula (Hymenoptera: Braconidae) to volatiles from different plant-herbivore complexes. J Insect Behav 9:525–538 Geervliet JBF, Posthumus MA, Vet LEM, Dicke M (1997) Comparative analysis of headspace volatiles from different caterpillar-infested or uninfested food plants of Pieris species. J Chem Ecol 23:2935–2954 Geervliet JBF, Vreugdenhil AI, Dicke M, Vet LEM (1998) Learning to discriminate between infochemicals from different plant-host complexes by the parasitoids Cotesia glomerata and C. rubecula. Entomol Exp Appl 86:241–252
123
Geervliet JBF, Verdel MSW, Snellen H, Schaub J, Dicke M, Vet LEM (2000) Coexistence and niche segregation by field populations of the parasitoids Cotesia glomerata and C. rubecula in the Netherlands: predicting field performance from laboratory data. Oecologia 124:55–63 Gols R, Witjes LMA, van Loon JJA, Posthumus MA, Dicke M, Harvey JA (2008) The effect of direct and indirect defenses in two wild brassicaceous plant species on a specialist herbivore and its gregarious endoparasitoid. Entomol Exp Appl 128:99–108 Gols R, Bullock JM, Dicke M, Bukovinszky T, Harvey JA (2011) Smelling the wood from the trees: non-linear parasitoid responses to volatile attractants. J Chem Ecol 37:795–807 Graves SD, Shapiro AM (2003) Exotics as host plants of the California butterfly fauna. Biol Conserv 110:413–433 Gripenberg S, Mayhew PJ, Parnell M, Roslin T (2010) A metaanalysis of preference–performance relationships in phytophagous insects. Ecol Lett 13:383–393 Harvey JA, Fortuna T (2012) Chemical and structural effects of invasive plants on herbivore-parasitoid/predator interactions in native communities. Entomol Exp Appl 144:14–26 Harvey JA, Biere A, Fortuna T, Vet LEM, Engelkes T, Morrien E, Gols R, Verhoeven K, Vogel H, Macel M, HeidelFischer HM, Schramm K, van der Putten WH (2010a) Ecological fits, mis-fits and lotteries involving insect herbivores on the invasive plant, Bunias orientalis. Biol Invasions 12:3045–3059 Harvey JA, Bukovinszky T, van der Putten WH (2010b) Interactions between invasive plants and insect herbivores: a plea for a multitrophic perspective. Biol Conserv 143:2251–2259 Hood GM (2010) PopTools version 3.2.3. Available on the internet. http://www.poptools.org Jaenike J (1978) Optimal oviposition behavior in phytophagous insects. Theor Popul Biol 14:350–356 Jaenike J (1990) Host specialization in phytophagous insects. Annu Rev Ecol Syst 21:243–273 Jervis MA, Ellers J, Harvey JA (2008) Resource acquisition, allocation, and utilization in parasitoid reproductive strategies. Annu Rev Entomol 53:361–385 Keane RM, Crawley MJ (2002) Exotic plant invasions and the enemy release hypothesis. Trends Ecol Evol 17:164–170 Keeler MS, Chew FS (2008) Escaping an evolutionary trap: preference and performance of a native insect on an exotic invasive host. Oecologia 156:559–568 Ku¨hnle A, Mu¨ller C (2009) Differing acceptance of familiar and unfamiliar plant species by an oligophagous beetle. Entomol Exp Appl 131:189–199 Larsson S, Ekbom B (1995) Oviposition mistakes in herbivorous insects: confusion or a step towards a new host giant. Oikos 72:155–160 Mayhew PJ (1997) Adaptive patterns of host-plant selection by phytophagous insects. Oikos 79:417–428 Mumm R, Dicke M (2010) Variation in natural plant products and the attraction of bodyguards involved in indirect plant defense. Can J Zool 88:628–667 Ohsaki N, Sato Y (1994) Food plant choice of Pieris butterflies as a trade-off between parasitoid avoidance and quality of plants. Ecology 75:59–68 Papaj DR (2000) Ovarian dynamics and host use. Annu Rev Entomol 45:423–448
A tritrophic approach to the preference–performance hypothesis Payne RW, Harding SA, Murray DA, Soutar DM, Baird DB, Glaser AI, Welham SJ, Gilmour AR, Thompson R, Webster R (2011) Guide to regression, nonlinear and generalized linear models in genstat. VSN International, Hemel Hempstead Pierre PS, Jansen JJ, Hordijk CA, van Dam NM, Cortesero AM, Dugravot S (2011) Differences in volatile profiles of turnip plants subjected to single and dual herbivory above- and belowground. J Chem Ecol 37:368–377 Price PW, Bouton CE, Gross P, Mcpheron BA, Thompson JN, Weis AE (1980) Interactions among 3 trophic levels: influence of plants on interactions between insect herbivores and natural enemies. Annu Rev Ecol Syst 11:41–65 Rausher MD (1979) Larval habitat suitability and oviposition preference in three related butterflies. Ecology 60:503–511 Renwick JAA (2002) The chemical world of crucivores: lures, treats and traps. Entomol Exp Appl 104:35–42 Scheirs J, De Bruyn L, Verhagen R (2003) Host nutritive quality and host plant choice in two grass miners: primary roles for primary compounds? J Chem Ecol 29:1373–1389 Shiojiri K, Takabayashi J, Yano S, Takafuji A (2002) Oviposition preferences of herbivores are affected by tritrophic interaction webs. Ecol Lett 5:186–192 Singer MC, Ng D, Thomas CD (1988) Heritability of oviposition preference and its relationship to offspring performance within a single insect population. Evolution 42:977– 985 Smid HM, van Loon JJA, Posthumus MA, Vet LEM (2002) GCEAG-analysis of volatiles from Brussels sprouts plants damaged by two species of Pieris caterpillars: olfactory receptive range of a specialist and a generalist parasitoid wasp species. Chemoecology 12:169–176 Stamp N (2001) Enemy-free space via host plant chemistry and dispersion: assessing the influence of tri-trophic interactions. Oecologia 128:153–163 Takabayashi J, Dicke M (1996) Plant-carnivore mutualism through herbivore-induced carnivore attractants. Trends Plant Sci 1:109–113
Thompson JN (1988) Evolutionary ecology of the relationship between oviposition preference and performance of offspring in phytophagous insects. Entomol Exp Appl 47: 3–14 Thompson JN, Pellmyr O (1991) Evolution of oviposition behavior and host preference in Lepidoptera. Annu Rev Entomol 36:65–89 Travers-Martin N, Mu¨ller C (2008) Matching plant defence syndromes with performance and preference of a specialist herbivore. Funct Ecol 22:1033–1043 Turlings TCJ, Tumlinson JH, Lewis WJ (1990) Exploitation of herbivore-induced plant odors by host-seeking parasitic wasps. Science 250:1251–1253 Valladares G, Lawton JH (1991) Host-plant selection in the holly leaf-miner: does mother know best? J Anim Ecol 60:227–240 van Wijk M, de Bruijn PJA, Sabelis MW (2011) Complex odor from plants under attack: herbivore’s enemies react to the whole, not its parts. Plos One 6:e21742 Vet LEM, Dicke M (1992) Ecology of infochemical use by natural enemies in a tritrophic context. Annu Rev Entomol 37:141–172 Vet LEM, Lewis WJ, Carde RT (1995) Parasitoid foraging and learning. In: Carde RT, Bell WJ (eds) Chemical ecology of insects 2. Chapman & Hall, New York, pp 65–101 Vet LEM, De Jong AG, Franchi E, Papaj DR (1998) The effect of complete versus incomplete information on odour discrimination in a parasitic wasp. Anim Behav 55:1271– 1279 Vinson SB (1976) Host selection by insect parasitoids. Annu Rev Entomol 21:109–133 Vos M, Berrocal SM, Karamaouna F, Hemerik L, Vet LEM (2001) Plant-mediated indirect effects and the persistence of parasitoid-herbivore communities. Ecol Lett 4:38–45 West-Eberhard MJ (1989) Phenotypic plasticity and the origins of diversity. Annu Rev Ecol Syst 20:249–278
123